This proposal is aimed at further developing and applying a general method that allows one to systematically add amino acids with novel physical, chemical and biological properties to the genetic codes of both prokaryotic and eukaryotic organisms. This methodology involves the generation of orthogonal tRNA/aminoacyl-tRNA synthetase (aaRS) pairs that uniquely recognize a noncoding codon and do not cross react with any of the endogenous tRNAs and aminoacyl-tRNA synthetases in the host organism. The amino acid specificity of the aminoacyl-tRNA synthetase is then modified such that it aminoacylates its cognate tRNA with only the desired unnatural amino acid and no endogenous amino acid. Using this method, a large number of structurally diverse unnatural amino acids have been incorporated efficiently and with high fidelity into proteins in response to nonsense and frameshift codons. We now propose to extend these studies in several directions. The first is focused on developing robust, user-friendly methods to genetically encode a large number of these unnatural amino acids in mammalian cells and Streptomyces. If successful this effort will significantly expand the scope and impact of this technology. The second effort focuses on the synthesis and in vitro and in vivo characterization of chemically defined antibody conjugates for the treatment of cardiovascular disease and acute myeloid leukemia. These latter efforts should not only pave the way for new or significantly improved therapeutics for major unmet medical needs, they should also provide a generalizable, straightforward approach to create similar classes of antibody conjugates for metabolic, cardiovascular and inflammatory diseases.